* The role of the board of directors (BOD) in the corporate governance process had received considerable research attention in recent years. Although the board is legally authorized to ratify monitor managerial decisions, critics have argued strongly that management generally dominates the board by its influence on the selection of outside directors, and by its control over the agenda of board meetings and the information provided to outside board members. Hence, the issue of BOD effectiveness is an empirical question.
A growing body of recent empirical research appears to support the notion that outside directors are important in monitoring managerial actions. However, only a limited number of studies focus directly on the issue of the BOD's impact on shareholders' wealth (Byrd and Hickman , Rosenstein and Wyatt , and Shivdasani ). This study contributes to the evolving empirical literature by focusing on the narrow, but yet unexplored, issue of the role of the BOD in management buyouts. It examines whether wealth gains in MBOs are affected by the composition of the board of directors (the proportion of independent outsiders), and if so, whether the BOD helps to mitigate the agency issues inherent in these transactions.
In a management buyout (MBO), an investor group that includes management makes a bid for all or part of a publicly held corporation. The nature of these transactions inherently provides for conflicts of interest. While management, as potential owners of the privatized firm, has the fiduciary obligation to obtain the highest price for shareholders, it also has an incentive to make the acquisition at the lowest possible price. This conflict is more pronounced when the entire firm is taken private, because the firm's top management is generally part of the buyout group. Although the courts and regulatory agencies play a role in helping to ensure that shareholder interests are served, the primary responsibility for maximizing shareholder wealth rests with those corporate directors who are independent of management.
While the CEO and other top executives control the board of directors under normal circumstances, the role of outside directors assumes greater importance when serious agency problems surface between inside managers and shareholders, for example, in response to financial distress, an incompetent CEO, or takeover bid (see Lorsch ). In going-private transactions, independent outside directors are specifically charged with protecting shareholder interests (see Lederman and Bryer ).
Although a considerable body of empirical research has evolved on the operating characteristics and shareholder wealth effects of management buyouts (see , ,  and ), the role of the BOD in such transactions has not been examined. Our empirical results indicate that for bids where the entire firm is to be taken private, the increase in shareholder wealth is substantially higher when the board is numerically dominated by independent outside directors than when insiders and nonindependent outsiders control the board. For transactions in which management takes only a segment (business unit) of the firm private, the results are inconclusive. A comparison of these results indicates that independent directors play an important role in going-private transactions, where serious agency problems exist between top managers and shareholders. However, independent directors appear to be less important in unit management buyouts, where top management often negotiates at arm's length with the purchasing group. Additionally, in going-private transactions, high levels of inside ownership, which mitigate the agency problem between managers and shareholders, have a positive impact on abnormal returns.
The paper is organized as follows: Section I discusses conflicts of interest for manager and the role of the board of directors, and then presents research questions. Section II describes the sample selection procedure and empirical methods. Section III presents empirical results for going-private (GP) bids and unit-management buyout (UMB) bids, respectively. Conclusions are drawn in Section IV.
I. Empirical Issues
This section details the management incentives in MBOs, the role of the independent directors in such transactions, testable hypotheses, and prior research.
A. Management's Incentives
In summarizing recent empirical work on leveraged buyouts, LBOs, Jensen and Ruback  state that:
... value increases accompanying LBOs stem from
management practice -- they do no come at the
expense of bondholders, employees, or reduced
This improved management practice is often predicated on retaining and properly motivating senior management teams from pre-buyout firms. For example, Kohlberg Kravis Roberts & Co.'s criteria for participating in a buyout includes a strong management team that is often given the opportunity to own 10 to 20% of the equity in the privatized firm . As for managerial motivation, Jensen  notes that in Kaplan's  study of LBOs, the average CEO receives $64 per $1,000 change in shareholder wealth, as compared to $3.25 per $1,000 for the average Forbes 1000 Company.
In MBOs, management teams can benefit from a lower bid in two ways. First, their out-of-pocket (or borrowing) costs of acquiring equity are reduced. Second, by minimizing the amount of debt needed to complete the transaction, they substantially improve the probability that the privatized firm will be successful. These potential opportunities for personal gain are strong incentives for managements to deviate from their duty to maximize shareholder wealth.
B. The Role of Independent Directors in MBO Transactions
Lederman and Bryer  note that the leading Delaware case delineating the board of director's obligations in a management buyout is Weinberger v. UOP. In reference to the shareholders' class action suit challenging the board's conduct in the proposed merger of UOP Inc., into The Signal Companies, they quote from the court's decision:
When directors of a Delaware corporation are on
both sides of a transaction, they are required to
demonstrate their utmost good faith and the most
scrupulous inherent fairness of the bargain. The
requirement of fairness is unflinching in its demand
that when one stands on both sides of the transaction,
he has the burden of establishing its entire fairness.
[Weinberger v. UOP, Inc. (457 A.2d 701, Del. Sup.
Lederman and Bryer  further state that in view of Weinberger's emphasis on the subject, the board's decision-making process should be structured to maximize the role of the independent directors:
If the directors who are members of the buyout group
constitute a minority of the board, the disinterested
majority can act for the board, with the buyout group
members abstaining from both the vote and
participation in the deliberations on the buyout. In a
majority of the board is not disinterested, the board can
delegate (the independent committee) the power to
review and/or negotiate the terms of the buyout.,
Generally, the members of the independent committee have the full protection of the "business judgment rule," which protects from judicial second-guessing a director who acts in good faith, is adequately informed, and has no personal interest in the transaction. The obligation to act in good faith requires that the committee act in the best interest of shareholders by obtaining the best price for shareholder, and promoting a "level playing field" for bidders other than the management group. This responsibility may place a great burden on independent directors. The management group, having intimate knowledge of the firm's operations, has a distinct advantage in formulating its bid. Hence, the independent directors must ensure that competing bidders have access to all relevant documents, and that management is not granted any special considerations once the bidding process has been set into motion.
The obligation to be adequately informed is deceptively onerous. The independent committee, whose members seldom have intimate knowledge of the firm's operations, must be certain that bids from any party are at least fair, if not generous. This requirement is complicated because courts have held that a bid of some arbitrary multiple of the pre-bid stock price is not sufficient evidence of fair value (see [19, p. 128]). Thus, the independent committee must have the expertise and managerial skill to retain investment advisors and legal counsel, if need be, and to conduct a fair auction, often under intense public scrutiny and severe time constraints.
C. Testable Hypotheses
From the foregoing discussion, it is reasonable to expect that in GP transactions, boards with a greater proportion of independent directors may be more likely to form independent committees that have the expertise and the familiarity with the companies' operations to maximize shareholder value. Thus, we hypothesize that the market's initial response to the announcement of a GP bid will be greater for firms with boards that are numerically dominated by independent directors than those that are not, and that the higher the proportion of independent directors on the board, the higher the abnormal returns.
With UMBs, it is less likely that senior managers on the board will have a personal stake in the transaction and therefore these transactions appear less likely to pose a serious conflict of interest. Although the proportion of independent directors on the board may affect shareholder returns in UMB transactions, we have no a priori expectation as to the magnitude or direction of the effect. In the case of UMBs, the inside managers on the board are generally not on the management team of the unit. For example, in our sample of 74 UMBs, in only two instances did unit managers serve on the corporation's board. Thus, in UMBs, the insiders on the board should generally view their interests as in alignment with those of shareholders. Management should have sufficient incentive to negotiate the best possible price for shareholders, and the agency conflicts should be far less serious.
D. Prior Research
In general, prior research has found that shareholders of firms going private benefit from the transaction. DeAngelo, DeAngelo, and Rice  computed average two-day increases in value for GPs of 22.27% and an increase of 30.40% over a 40-day interval surrounding the announcement. Marais, Schipper, and Smith  reported average abnormal common stock returns of 13% over the announcement period with an additional 9% during the pre-announcement period. Other researchers (for example, Lehn and Poulsen , and Torabzadeh and Bertin ) reported similar findings.
UMBs are essentially sell-offs to business unit managers, involving only a portion of the firm's assets. Hite and Vetsuypens  found significant abnormal returns of 0.55% over a two-day announcement period for a sample of UMBs, while Trifts, Sicherman, Roenfeldt, and de Cossio  reported abnormal returns of 1%. Shareholders, on average, appear to benefit from UMBs, but as might be expected, abnormal returns are considerably smaller than those in GPs.
A growing body of empirical research supports the notion that outside directors are important in monitoring managerial actions. For example, Rosenstein and Wyatt  found that even though most boards are numerically dominated by outsiders, the addition of an outside director is associated with positive abnormal returns. Kosnik  found that greater diversity in outside directors' principal occupations increases the tendency of corporations to use greenmail, suggesting that diversity fragments the board and provides insiders with a greater degree of control. Both Cochran, Wood, and Jones , and Singh and Harianto  reported that firms adopting golden parachutes have a higher proportion of outside directors than those that do not, indicating that the protection afforded by a golden parachute more closely aligns management and shareholder interests. Weisbach  found that CEO turnover is more highly correlated with firm performance in corporations having a majority of outside directors than in those with a majority of insiders. Hermalin and Weisbach  found that outsiders are more likely to join a board after a firm performs poorly or leaves an industry.
With respect to corporate takeovers, Brickley and James  found that the proportion of outside directors is significantly lower on boards of banks in states that restrict banking acquisitions than in those that do not, suggesting that outside directors play a major role in evaluation takeover proposals. They also noted that in "nonacquisition" states, where there is no corporate control market to discipline managers, outsider dominated boards help control managerial consumption of perquisites (as proxied by salary expenses). Byrd and Hickman , in analyzing tender offers, reported that bidding firms with a majority of independent directors on the board earn significantly higher abnormal returns than those that do not.
Shivdasani  found for companies that are hostile takeover targets, and thus, are presumably not being managed efficiently, outside directors hold significantly lower ownership stakes than in firms that are not hostile takeover candidates. Outside directors in hostile takeover targets also hold fewer other directorships, indicating that they are less experienced or less effective monitors of management.
II. Sample Selection and Methodology
We use event study methodology to determine abnormal performance around the time of the announcement of an MBO. The first announcement of the MBO in the Wall Street Journal is day 0 in the event tests. We then use cross-sectional regressions to examine the relationships between abnormal performance and board composition, along with a variety of control variables.
A. Sample Selection
The initial samples of completed going-private transactions (GPs) and unit management buyouts (UMBs) were identified from Mergerstat Review  over the years 1983 to 1989. The sample was restricted to those firms listed on the New York and American Stock Exchanges (NYSE and AMEX) having stock returns on the CRSP Daily Stock Returns database over the period from 300 trading days before the first announcement of the MBO to 20 trading days after the announcement. To minimize the potential effects of event contamination, we eliminated firms with contemporaneous announcements of other events over the days -2 to +1 relative to the announcement.
Several additional constraints were imposed to mitigate any extraneous effects and also to enhance the quality of the analyses. Proxy statements for the regular annual meeting preceding the announcement had to be available from Q Data Corporation's microfiche file for all NYSE and AMEX sample firms. For each firm, we also examined the proxy statement prior to the one identified for analysis to determine if any of the firms had major changes in their board composition in the year before the announcement. Most firms had no changes in their boards although some had one addition or replacement. We started out with 88 GPs and 94 UMBs, but due to lack of information in the Wall Street Journal Index, unavailability of daily returns, market value data, or board composition, we were forced to eliminate 30 GPs and 20 UMBs. Our final samples consisted of 58 GPs and 74 UMBs. The 74 UMBs came from 58 firms, where some firms had multiple UMBs. The distribution of announcements by year is shown in Panel A of Exhibit 2 for GPs and Panel A of Exhibit 3 for UMBs. The firms in the GP sample come from 48 different three-digit SIC code industries, and for the UMB sample they come from 38, providing for considerable industry diversity.
B. Board Composition, Takeover Speculation, and Sample Characteristics
A number of researchers (, , , , and ) have noted that the traditional distinction between inside and outside directors fails to account for the realized and potential conflicts of interest between outside directors and the corporations they serve. Examples include: outside attorneys who perform legal work for the firm, officers of other corporations whose firms engage in trade with the firm, creditors, relatives of officers, etc. Additionally, in the case of outside directors from financial institutions, Brickley, Lease and Smith  note that some financial institution (e.g., insurance companies, banks and non-bank trusts) are more sensitive to pressure by inside managers than others (e.g., public pension funds, mutual funds, endowments, and foundations). Hence, in this study, we classify directors as insiders, affiliated outsiders, and independent outsiders according to Exhibit 1. These classifications closely follow Gilson  with several exceptions: Investment bankers are always classified as affiliated outsiders, whether or not they have performed services for the firm over the past year, and managers of other companies are classified as affiliated outsiders if the firms engage in substantial trade.(1)
Board composition statistics for GPs are shown in Panel B of Exhibit 2. Insiders represent 34.5% of the average board in our GP sample. Overall board size and the proportions of insiders and affiliated outsiders and similar to previous studies (see , ,  and ). Although over 93% of the board are outsider-dominated using the traditional insider-outsider classification, as shown in Panel C, this percentage drops to 81% (summing the 0 to 40% and 40 to 50% groups) when affiliated outsiders are classified with insiders.
Panel D presents the market value of common equity for the sample firms as well as dollar values of shares controlled by insiders, affiliated, outsiders, and independent outsiders. For consistency, market values are computed by multiplying the number of shares outstanding or controlled by each class of directors by the share price on the last day of the month that ends at least thirty days before the announcement. Comparing the sample means and medians reveals that although the means ownership by inside and outside directors is 3.4% and 1.5%, respectively, of the firms' market values, the distribution is skewed by extremely large shareholders in just a few firms. Outside directors usually own just nominal amounts of stock, but in several firms, large shareholders and trustees have board seats.
Firms that are the subject of takeover speculation might be expected to have lower abnormal returns upon the GP announcement, because their pre-announcement prices already reflect expectations regarding the bid. Fourteen firms is our sample were subjects of takeover speculation or activity one year prior to the formal GP announcement as per the Wall Street Journal. For nine of those 14 firms in play, management initiated the GP transaction in response to a takeover bid; we classify these nine announcements as "defensive."
Since firms may have been "in play" for quite some time before the GP announcement, stock performance over a prolonged period of time may have a bearing on announcement period returns. We define abnormal prior performance as the sum of abnormal returns over the 130-trading-day period from 150 to 21 trading days before the announcement (roughly six calendar months), based on market model parameters estimated over the period from 300 to 151 trading days before the announcement. Panel E shows that, on average, prior abnormal performance is roughly 0.11, and statistically greater than zero at the 0.05 level. However, abnormal performance is extreme, ranging from -0.89 to 0.98, and similar to that reported by Marais, Schipper, and Smith .(2)
In Exhibit 3, we present the board and firm characteristics for the UMB sample. Panel B shows that insiders average 32.8% of the total board with affiliated outsiders averaging 6.9%. Panel C shows that 73 of the 74 boards (98.6%) have outsiders comprising at least 50% of the board, but this number drops to 65 when affiliated outsiders are classified with insiders. Panel D shows the total market value of equity for each firm and the market value of equity owned by various categories of board members.
In our UMB sample, 19 of the 74 firms were in play during the preceding 12 months, but only 4 UMBs appeared to be defensive. Panel E contains the abnormal returns computed for the interval -150 to -21 days, which, on average, is 0.0615. To give an impression of the size of the unit being taken private relative to the size of the firm in question, Panel F shows the ratio of the bid for the assets to the market value of the firm. This statistic averaged 21.23%, but ranged from 0.6% to 323.0%.
C. Statistical Methods
We employ standard event study methodology to generate abnormal returns using a 150-trading-day period from days -300 to -151 relative to the announcement to estimate market model parameters. Returns preceding day -150 are used in generating market model parameters as returns over the period from days-150 to -21 relative to the announcements are used to obtain independent estimates of prior firm performance.(3)
Average abnormal returns and cumulative abnormal returns (CARs) are computed over various intervals within the period -20 to +20. Statistical tests used to measure abnormal performance are similar to those in Dodd and Warner . The two-sample comparison Z-statistic used to test for differences between CARs of subsamples is as follows: (1) [Mathematical Expression Omitted] where [Z.sub.1] and [Z.sub.2] are the Z-statistics for the individual subsamples, and [n.sub.1] and [n.sub.2] are the respective subsamples sizes.(4)
We use cross-sectional regressions to test hypotheses regarding the effects of board composition on abnormal performance, while simultaneously controlling for a number of other variables. Cross-sectional also accounts for any variance increases over the test period, in contrast to time-series estimation period variances.
Continuous and dummy variables are used in various specifications to denote the proportions of either affiliated plus independent outsiders or only independent outsiders on the board. These variables are defined as follows:
Pout = proportion of independent + affiliated outside directors on the board;
Outdom = 1 if the proportion of independent and
affiliated outside directors is greater than 0.5; 0 otherwise; Pind = proportion of independent outside directors on the board; and
Inddom = 1 if the proportion of independent outside
directors is 0.5 or larger; 0 otherwise.
Abnormal returns may also be related to a number of factors other than BOD composition, such as firm size, whether or not the firm was subject to takeover activity before the announcement, performance before the announcement, and share ownership levels by both inside and outside directors. Variables created to control for these factors are defined as follows:
Pinown = the proportion of share ownership by
insiders on the board relative to the total number of shares outstanding;
Ownrat = the ratio of share ownership by insiders and
affiliated outsiders relative to ownership by independent outsiders;
Lmvrel = the natural logarithm of the ratio of the
market value of common equity for the firm in question to the market value of the average firm in the sample;
Inplay = 1 if any story regarding takeover appeared
in the Wall Street Journal in the year before the announcement; 0 otherwise; Def = 1 if the buyout occurs subsequent to a takeover bid; 0 otherwise;
Pperf = the sum of prediction errors over the
130-trading-day period from day -150 to -21 relative to the announcement; and
Usize = for UMBs, the ratio of the bid for the unit
relative to the market value of the entire firm.
The dependent variable is the CAR for each firm over the two-day interval including days -1 and 0.(5) Weighted least square regression is used to account for the heteroscedasticity in abnormal returns across firms. The variables for each observation are weighted by the reciprocal of the standard error of the time-series regression used to estimate market parameters.(6)
A. CAR Results -- GP Transactions
Panel A of Exhibit 4 contains the CAR results. Using a delineation of inside/outside director, both subsamples have significantly positive CARs. Boards dominated by insiders have a CAR for the interval -1 to 0.1036 (Z = 10.82), while boards dominated by outsiders have a CAR of 0.1576 (Z = 39.85). The difference -0.0540 (Z = -4.75) is significant with the parametric Z-tests, but insignificant using the Wilcoxon nonparametric test. Ninety-two percent of boards dominated by outsiders have positive two-day CARs as compared to 37.5% for insider dominated boards. While the magnitude of the numbers varies, the results are generally consistent for the intervals -5 to 0 and -10 to 0, but over the interval -20 to 0 the difference between the two subsamples is significant at conventional levels of significance.
In Panel B of Exhibit 4, we have recomputed the results by dividing the samples into boards dominated by nonindependent directors (insiders and affiliated outsiders) and those dominated by independents. The results demonstrate that boards dominated by independent members earn larger abnormal returns than those dominated by nonindependent members. Over the interval -1 to 0, the CAR for nonindependent boards is 0.0295 (Z = 3.44), but for independent dominated boards it is 0.1784 (Z = 43.91). The difference, -0.1489, is statistically significant using both the Z-test and Wilcoxon test. In addition, 95.7% of the CARs are positive for independent boards in comparison to 36.4% for the nonindependent boards. Similar results are observed over the intervals -5 to 0, -10 to 0, and -20 to 0. Consistent with our hypothesis, the market's initial response to announcement of a going-private bid is greater for those firms dominated by independent directors. This difference is more pronounced when independent directors are classified as nonaffiliated outsiders than when they are grouped together with affiliated outsiders. In addition, overall abnormal returns are reasonably close to those reported in prior research (see Kaplan  and Torabzadeh and Bertin ).
B. Cross-Sectional Results -- GP Transactions
Exhibit 5 details the cross-sectional regression results for the GP transactions with the CAR from -1 to 0 as the dependent variable. In regression 1, the variable Pout, which measures the proportion of outsiders (independent plus affiliated outsiders) on a board, is significantly positive, as predicted by our hypothesis. The larger the proportion of outsiders, the more positive the abnormal returns when a GP bid is announced. A higher ratio of outsiders on the board appears to mitigate the agency problem inherent in a GP transaction. Similarly, the control variable Pinown, the proportion of ownership by insiders, is also significant and positively related to the announcement period abnormal returns, implying that higher share ownership by insiders helps in alleviating the agency problem by providing the inside board members incentives to align themselves with shareholder interests. None of the remaining control variables are significant, although the signs of the coefficients are as we would expect.(7)
Regression 2 is similar to regression 1, except that we use a dummy variable that is 1 if a board is dominated by outsiders and zero otherwise (Outdom), instead of percentage of outside control. As expected, this variable is positively related to the announcement period abnormal returns.
Regressions 3 and 4 differentiate board control using the alternative definition of board member affiliation. In regression 3, the variable Pind, which measures the percentage of independent outsiders, is significantly positive, as predicted by our hypothesis. In regression 4, the variable Inddom is a dummy variable with the value of 1 when the proportion of independent outside directors is 50% or larger. This variable is significant at better than 0.01. When a board is dominated by independent outside directors, the abnormal returns associated with a GP announcement are more positive than when boards are dominated by nonindependent members. This regression has the largest F-statistic (F = 2.66; significant at 0.02) and the largest [R.sup.2] (27%) of any of the regression in Exhibit 5.(8)
Exhibit 6 presents the board composition coefficients for regressions where the dependent variables were abnormal returns for the intervals -5 to 0, -10 to 0, and -20 to 0. The same control variables were included, but were insignificant in all regression. The results for these longer intervals are less conclusive because the cross-sectional variance in returns over longer intervals is larger, yielding less powerful results.
In general, our results are consistent with the hypothesis that the market's initial response to the announcement of a GP transaction is greater for firms whose boards are numerically dominated by independent directors. This relationship tends to be stronger when we classify affiliated outsiders with insiders and consider them both to be nonindependent. In addition, the level of inside director share ownership appears to be of some importance in determining abnormal returns.
C. UMB Results
Exhibit 7 contains our CAR results for the UMB transactions. Panel A shows the UMB results comparing those boards dominated by insiders with those dominated by outsiders. The CAR over the interval -1 to 0 is 0.0098 (Z = 0.95) for UMB announcements by firms with insider dominated boards. For those with outsiders dominated boards, the CAR is 0.0128 (Z = 5.05), which is statistically significant. The difference between these CARs has the predicted sign, but this difference is statistically insignificant with both the Z and Wilcoxon tests. Results over the other reported intervals, -5 to 0, -10 to 0, and -20 to 0 contain somewhat similar results. These results contain abnormal returns similar in size to those found in prior research (see Hite and Vetsuypens , and Trifts, et al ).
Panel B of Exhibit 7 details the results when we classify boards by nonindependent domination (insiders plus affiliated outsiders) versus independent domination (unaffiliated outsiders). Over the interval -1 to 0, the nonindependent dominated boards have a CAR which is insignificantly negative (CAR = -0.0042, Z = -0.61). For independent dominated boards, the CAR is 0.0149 (Z = 5.66). The difference, -0.0191, is significant using both the Z and Wilcoxon tests. These tests lend support to the hypothesis that for UMB announcements, abnormal returns are larger when the boards are dominated by independent members. Over other intervals, -5 to 0, -10 to 0, and -20 to 0, this difference has the same sign (CARs for independent boards larger than those for nonindependent boards), but in these intervals, the difference falls considerably short of significance.
We also tested the UMB results with cross-sectional regressions similar to those tests conducted on GPs, except that we add a variable to account for the size of the unit relative to the size of the whole firm (U size). For UMBs, the regression results appear in Exhibit 8. None of the coefficients measuring board composition are significant and the nominal signs of these coefficients vary unpredictably. While we found significance in our time-series Z-test (reported in Exhibit 7), these differences do not remain in the cross-sectional regressions where we control for other variables.(9) The coefficient for the variable Lmvrel is negative and strongly significant, indicating that abnormal returns are inversely related to the market value of the firm being examined, although the coefficient for the variable measuring the relative size of the unit being divested (Usize) is insignificant. This somewhat surprising size effect is difficult to rationalize. One potential explanation is that large firms may signal future divestiture programs more effectively, resulting in less new information associated with specific announcements.
Exhibit 9 present the board composition coefficients for regressions where the dependent variables were abnormal returns for the intervals -5 to 0, -10 to 0, and -20 to 0. The same control variables were included and generally yielded qualitatively similar results to those for the interval -1 to 0.
D. Comparison of GP and UMB Results
We found that boards dominated by independent directors earned larger returns when announcing a GP transaction than do boards dominated by nonindependents.(10) This difference occurred consistently across our definitions of independence and variable design. Yet, for UMBs, the influence of the board composition is not nearly as strong. The differences may be attributed to those differences in agency costs between GPs and UMBs. Senior management on boards do not generally have a personal stake in the UMB transactions, but would generally benefit by purchasing assets at a low price (at the expense of shareholders) in GPs. Alternatively, our tests may lack the power to discern differences in the smaller abnormal returns associated with UMBs.
In this study, we note that there is an agency problem in management buyouts because participating managers may have conflicting interests with shareholders. We hypothesize that members of boards who are independent of the firm should play a role in reducing this conflict. As anticipated, the results indicate that boards dominated by independent members are associated with larger abnormal returns in transactions in which management participates in taking the entire firm private. Our results are consistent with the point of view that when inside board members own larger percentages of common stock in the firm, this agency problem appears to be further mitigated, as evidenced by its positive relationship to abnormal returns. While there is some very weak evidence to support the conclusion that independent board members positively influence shareholder returns in UMB transactions, strong conclusions cannot be made. Our results suggest that the role of the board of directors in UMBs may be less important in addressing any agency problems or that agency problems in UMBs may be less pronounced. (1)Proxy statements disclose the amount of transactions with directors and their firms over the past year, but in the case of investment bankers they do not necessarily disclose an ongoing relationship. (2)Although negative absolute returns are bounded at -1.00, negative abnormal returns may be less than -1.00. (3)Event study results are essentially unchanged the 150-trading-day period from days -170 to -21 relative to the announcement for parameter estimation. (4)Since the individual abnormal returns are standardized for the tests, it is assumed the variance of the distribution of abnormal returns for each security is unity. (5)Regression with the CAR over longer intervals (i.e., -5 to 0, -10 to 0, -20 to 0) were also run to account for information leakage before the announcement. (6)Tests of significance for all specifications are essentially unchanged when ordinary least squares is used, or when the standardized CAR is used as the dependent variables in OLS regressions. (7)The control variables generally have the sign that we would predict, although their coefficients are generally insignificant. The variable In-play shows that the announcement period returns are smaller when previous bids were made. Previous offers may have already inflated the stock price. Defensive GP bids have a positive coefficient, which may reflect a board's attempt to raise bid prices, which is consistent with shareholder wealth maximization. The variable Pperf measures prior performance (CAR over -150 to -21), and it is negatively related to the abnormal returns. Stock price increases to previous bids may explain this coefficient's sign as well as possible speculation over the current bid. (8)Overall F-statistics were higher when some of the insignificant variables were removed from the regressions, but we believe that all of these variables are required to properly specify the model. (9)The control variables in the UMB regressions show that defensive UMBs earn significantly larger returns than nondefensive. The variable, Lmvrel, is significantly negative in all regressions. Larger firms that announce UMBs earn smaller returns. Percentage ownership of stock by insiders has no effect on the results for UMBs. (10)We tested our results to determine if there was a time-series trend in the abnormal returns. In our sample, we found no significant differences for returns across years and no evidence of a trend.
B. Baysinger and H. Butler, "Corporate Governance and the Board of Directors: Performance Effects of Changes in Board Composition," Journal of Law, Economics, and Organization (Fall 1985), pp. 101-124. J.A. Brickley and C.M. James, "The Takeover Market, Corporate Board Composition, and Ownership Structure: The Case of Banking," Journal of Law and Economics (April 1987), pp. 161-180. J.A. Brickley, R.C. Lease, and C.W. Smith, Jr., "Ownership Structure and Voting on Antitakeover Amendments," Journal of Financial Economics (January/March 1988), pp. 267-291. J. Byrd and K. Hickman, "Board of Director Composition and the Profitability and Shareholder Wealth Effects of Acquisition," Working Paper, Washington State University, 1990. P.L. Cochran, R.A. Wood, and T.B. Jones, "The Composition of Boards of Directors and Incidence of Golden Parachutes," Academy of Management Journal (September 1985), pp. 664-671. H. DeAngelo, L. DeAngelo, and E.M. Rice, "Going Private: The Effects of a Change in Corporate Ownership," Midland Corporate Finance Journal (Summer 1984), pp. 35-43. P. Dodd and J.B. Warner, "On Corporate Governance: A Study of Proxy Contests," Journal of Financial Economics (April 1983), pp. 401-438. E.F. Fama and M.C. Jensen, "Separation of Ownership and Control," Journal of Law and Economics (June 1983), pp. 301-325. S.C. Gilson, "Bankruptcy, Boards, Banks and Blockholders: Evidence on Changes in Corporate Ownership and Control When Firms Default," Journal of Financial Economics (October 1990), pp. 355-387. B.E. Hermalin and M.S. Weisbach, "The Determinants of Board Composition," RAND Journal of Economics (Winter 1988), pp. 95-112. G.L. Hite and M. Vetsuypens, "Management Buyouts of Divisions and Shareholder Wealth," Journal of Finance (September 1989), pp. 953-970. M.C. Jensen, "Active Investors, LBOs and the Privatization of Bankruptcy," Journal of Applied Corporate Finance (Spring 1989), pp. 35-44. M.C. Jensen and W.H. Meckling, "Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure," Journal of Financial Economics (October 1976), pp. 305-360. M.C. Jensen and R.S Ruback, "Preface," Journal of Financial V
Economics (September 1990), pp. 1-3. S. Kaplan, "The Effects of Management Buyouts on Operating Performance and Value," Journal of Financial Economics (October 1989), pp. 217-254. S.N. Kaplan and J.C. Stein, "How Risky is the Debt in Highly Leveraged Transactions?," Journal of Financial Economics (September 1990), pp. 215-246. Kohlberg Kravis and Roberts (with Deloitte Haskins and Sells), "Leveraged Buyouts," Journal of Applied Corporate Finance (Spring 1989), pp. 64-70. R.H. Kosnik, "Effects of Board Demography and Directors' Incentives on Corporate Greenmail Decisions," Academy of Management Journal (March 1990), pp. 129-150. L. Lederman and B.A. Bryer, "Representing a Public Company in a Leveraged Buyout Transaction," in Leveraged Management Buyouts: Causes and Consequences, Y. Amihud (ed.), Homewood, IL, Dow Jones-Irwin, 1989. K. Lehn and A.B. Poulsen, "Sources of Value in Leveraged Buyouts," Working Paper, University of Georgia, 1987. W.B. Little and S.B. Klinsky, "How Leveraged Buyouts Can Really Work: A Look at Three Cases," Journal of Applied Corporate Finance (Spring 1989), pp. 71-75. J.W. Lorsch with E. MacIver, Pawns or Potentates: The Reality of America's Corporate Boards, Boston, MA, Harvard Business School Press, 1989. L. Marais, K. Schipper, and A. Smith, "Wealth Effects of Going Private for Senior Securities," Journal of Financial Economics (June 1989), pp. 155-191. Merrill Lynch Business Brokerage and Valuation, Mergerstat Review, Schaumberg, IL, 1989. S. Rosenstein and J.G. Wyatt, "Outside Directors, Board Independence, and Shareholder Wealth," Journal of Financial Economics (August 1990), pp. 175-192. A. Shivdasani, "Board Composition, Ownership Structure, and Hostile Takeovers," Working Paper, Ohio State University, 1991. H. Singh and F. Harianto, "Management-Board Relationships, Takeover Risk, and the Adoption of Golden Parachutes," Academy of Management Journal (March 1989), pp. 7-24. K.M. Torabzadeh and W.J. Bertin, "Leveraged Buyouts and Shareholder Returns," Journal of Financial Research (Winter 1987), pp. 313-319. J.W. Trifts, N.W. Sicherman, R.L. Roenfeldt, and F. de Cossio, "Divesture to Unit Managers and Shareholder Wealth," Journal of Financial Research (Summer 1990), pp. 167-172. M.S. Weisbach, "Outside Directors and CEO Turnover," Journal of Financial Economics (January/March 1988), pp. 431-460. Chun I. Lee is a Doctoral Student, Stuart Rosentein is an Assistant Professor, Nanda Rangan is an Associate Professor, and Wallace N. Davidson III is Hehn Professor all with the Finance Department at Southern Illinois University, Carbondale, Illinois.…